# Just a Scratch: Enhancing LLM Capabilities for Self-harm Detection through Intent Differentiation and Emoji Interpretation

arXiv: https://arxiv.org/abs/2506.05073 · HTML: https://mlsanity.com/paper/2506.05073/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- Builds on 9 of its 37 references that are in the map.
- First posted on arXiv 5 Jun 2025.

## Builds on (most cited first, 9 of 9)

- [SQuAD: 100,000+ Questions for Machine Comprehension of Text](https://mlsanity.com/paper/1606.05250/) (2016)
- [LLM2Vec: Large Language Models Are Secretly Powerful Text Encoders](https://mlsanity.com/paper/2404.05961/) (2024)
- [Recognizing Emotion Cause in Conversations](https://mlsanity.com/paper/2012.11820/) (2020)
- [Training Models to Generate, Recognize, and Reframe Unhelpful Thoughts](https://mlsanity.com/paper/2307.02768/) (2023)
- [Risks from Language Models for Automated Mental Healthcare: Ethics and Structure for Implementation](https://mlsanity.com/paper/2406.11852/) (2024)
- [Looking Right is Sometimes Right: Investigating the Capabilities of Decoder-only LLMs for Sequence Labeling](https://mlsanity.com/paper/2401.14556/) (2024)
- [Boosting Distress Support Dialogue Responses with Motivational Interviewing Strategy](https://mlsanity.com/paper/2305.10195/) (2023)
- [Revisiting Instruction Fine-tuned Model Evaluation to Guide Industrial Applications](https://mlsanity.com/paper/2310.14103/) (2023)
- [Testing the Limits of Unified Sequence to Sequence LLM Pretraining on Diverse Table Data Tasks](https://mlsanity.com/paper/2310.00789/) (2023)

## Built on by (most cited first, 7 of 7)

- [CRADLE Bench: A Clinician-Annotated Benchmark for Multi-Faceted Mental Health Crisis and Safety Risk Detection](https://mlsanity.com/paper/2510.23845/) (2025)
- [Assessing Risks of Large Language Models in Mental Health Support: A Framework for Automated Clinical AI Red Teaming](https://mlsanity.com/paper/2602.19948/) (2026)
- [Responsible Evaluation of AI for Mental Health](https://mlsanity.com/paper/2602.00065/) (2026)
- [Two-Stage Voting for Robust and Efficient Suicide Risk Detection on Social Media](https://mlsanity.com/paper/2510.08365/) (2025)
- [Expert-Level Crisis Detection in Mental Health Conversations](https://mlsanity.com/paper/2606.10380/) (2026)
- [Analysing Self-Harm Representations in Language Models: a Cross-Architecture Study](https://mlsanity.com/paper/2607.21988/) (2026)
- [From Pattern Recognizers to Personalized Companions: A Survey of Large Language Models in Mental Health](https://mlsanity.com/paper/2609.25186/) (2026)

## Abstract (arXiv)

Self-harm detection on social media is critical for early intervention and mental health support, yet remains challenging due to the subtle, context-dependent nature of such expressions. Identifying self-harm intent aids suicide prevention by enabling timely responses, but current large language models (LLMs) struggle to interpret implicit cues in casual language and emojis. This work enhances LLMs' comprehension of self-harm by distinguishing intent through nuanced language-emoji interplay. We present the Centennial Emoji Sensitivity Matrix (CESM-100), a curated set of 100 emojis with contextual self-harm interpretations and the Self-Harm Identification aNd intent Extraction with Supportive emoji sensitivity (SHINES) dataset, offering detailed annotations for self-harm labels, casual mentions (CMs), and serious intents (SIs). Our unified framework: a) enriches inputs using CESM-100; b) fine-tunes LLMs for multi-task learning: self-harm detection (primary) and CM/SI span detection (auxiliary); c) generates explainable rationales for self-harm predictions. We evaluate the framework on three state-of-the-art LLMs-Llama 3, Mental-Alpaca, and MentalLlama, across zero-shot, few-shot, and fine-tuned scenarios. By coupling intent differentiation with contextual cues, our approach commendably enhances LLM performance in both detection and explanation tasks, effectively addressing the inherent ambiguity in self-harm signals. The SHINES dataset, CESM-100 and codebase are publicly available at: https://www.iitp.ac.in/~ai-nlp-ml/resources.html#SHINES .